Speech recognition system and method

• US 6,125,341 A
• Filed: 12/19/1997
• Issued: 09/26/2000
• Est. Priority Date: 12/19/1997
• Status: Expired due to Term

• Current Assignee: RPX Clearinghouse LLC (RPX Corporation)
• Original Assignee: Nortel Networks Corporation
• Inventors: Brennan, Paul M., Raud, Henry F.
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): Lerner, Martin
• Application Number:
  US08/994,450
• Time in Patent Office: 1,012 Days
• Field of Search: 704/231, 704/233, 704/243, 704/244, 704/246, 704/251, 704/255, 704/256, 704/257, 704/250, 704/270, 704/275, 704/3, 704/8, 704/10, 379/88.05, 379/88.06
• US Class Current: 704/8
• CPC Class Codes: G10L 15/005 Language recognition
• CPC Class Codes: G10L 15/187 Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 15/32 Multiple recognisers used in sequence or in parallel; Score combination systems therefor, e.g. voting systems
• CPC Class Codes: G10L 2015/228 of application context

Abstract

A speech recognition system having multiple recognition vocabularies, and a method of selecting an optimal working vocabulary used by the system are disclosed. Each vocabulary is particulary suited for recognizing speech in a particular language, or with a particular accent or dialect. The system prompts a speaker for an initial spoken response; receives the initial spoken response; compares the response to each of a set of possible responses in an initial speech recognition vocabulary to determine a response best matched in the initial vocabulary. A working speech recognition vocabulary is selected from a plurality of speech recognition vocabularies, based on the best matched response.

Claims

1. A method of recognizing speech at a speech recognition system using a plurality of speech recognition vocabularies, comprising:

selecting a working speech recognition vocabulary for recognizing speech from said plurality of speech recognition vocabularies based on a response by a user to an initial prompt;

receiving a spoken word to be recognized;

responding to determining a low likelihood of correctly recognizing said spoken word using said working speech recognition vocabulary by selecting another one of said plurality of speech recognition vocabularies, based on a response to a subsequent prompt, for use as said working speech recognition vocabulary;

said method further comprising presenting said user an option of receiving future prompts in a second language different from a language of said initial prompt based on a response to at least one of said initial prompt and said subsequent prompt.

2. The method of claim 1, wherein said response to said initial prompt is spoken by said user and compared to an initial speech recognition vocabulary to select said working speech recognition vocabulary.

3. The method of claim 2, wherein said response to said subsequent prompt is spoken by said user and compared to a re-evaluation vocabulary different from said initial speech recognition vocabulary to select said another one of said plurality of speech recognition vocabularies.

4. The method of claim 3, wherein said option is one of a plurality of options presented to said user based on at least one of a best match to said response to said initial prompt in said initial speech recognition vocabulary and a best match to said response to said subsequent prompt in said re-evaluation vocabulary.

5. The method of claim 1, wherein said second language is related to said working speech recognition vocabulary.

6. The method of claim 1, wherein said speech recognition system is a speaker independent speech recognition system.

7. The method of claim 6, wherein each of said plurality of vocabularies is optimized for recognition of speech having a particular dialect or accent.

8. The method of claim 1, wherein said subsequent prompt prompts said user for input used to choose said another one of said plurality of speech recognition vocabularies.

9. The method of claim 1, wherein said initial and subsequent prompts are audibly presented to said user.

10. The method of claim 1, wherein said initial and subsequent prompts are audibly presented to said user by way of a telephone network.

11. A speech recognition system, comprising

a processor;

an input device in communication with said processor;

a memory device in communication with said processor, said memory device storing data corresponding to a plurality of distinct speech recognition vocabularies;

said processor adapted to;

select a working speech recognition vocabulary for recognizing speech from said plurality of speech recognition vocabularies based on a response by a user to an initial prompt;

receive data representative of a spoken word to be recognized;

respond to determining a low likelihood of correctly recognizing said spoken word using said working speech recognition vocabulary by selecting another one of said plurality of speech recognition vocabularies, based on a response to a subsequent prompt, for use as said working speech recognition vocabulary;

said processor further adapted to present said user an option of receiving future prompts in a second language different from a language of said initial prompt based on a response to at least one of said initial prompt and said subsequent prompt.

12. The system of claim 11, further comprising a digital to analog converter to audibly present said initial and subsequent prompts to said user.

13. The system of claim 11, further comprising an analog to digital converter for receiving said data from an analog signal.

14. The system of claim 13, wherein said system is a speaker independent speech recognition system.

15. The system of claim 14, wherein each of said plurality of vocabularies is optimized for recognition of speech having a particular dialect or accent.

16. The system of claim 11, further comprising an interface to the public telephone network, to receive said spoken word by way of said public telephone network.

17. A method of recognizing speech at a speech recognition system using a plurality of speech recognition vocabularies, said method comprising:

receiving a response to an initial prompt from a user, said response used to select one of said plurality of speech recognition vocabularies to recognize speech;

based on said response, presenting said user an option of receiving subsequent prompts in a language different from said initial prompt.